Timeline for How to express a Poisson regression as an equation
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 28, 2014 at 23:57 | comment | added | Glen_b | rama - for offsets the coefficient is fixed at 1 - that's what offsets are designed to do. | |
| May 28, 2014 at 16:16 | comment | added | tchakravarty | @rama That is the point of offset terms. There are a number of motivations for how they come about, and you might want to read about them here and here. | |
| May 28, 2014 at 15:59 | comment | added | rama | Right, but this is where it gets confusing to me: if the offset is a model regressor, why is there no beta parameter estimate for it? | |
| May 28, 2014 at 15:54 | comment | added | tchakravarty | @rama While there is no problem with that, and what I have written is more general, the model should make substantive sense -- that the offset variable is also a regressor of the model. | |
| May 28, 2014 at 15:39 | comment | added | rama | Great, now I understand the form for the factor variable. Now, do the subscripts for the x covariate in the offset and B2 terms indicate these are two different variables? In my model, the x in P(x) is the same value as the x in the B2 term (as you've expressed it). | |
| May 28, 2014 at 15:29 | history | answered | tchakravarty | CC BY-SA 3.0 |